Lead, innovate, and grow with AI and generative AI

So we're going to get started now for our lead Innovate and Grow with AI and Generative AI Webinar. Today. I've got with us Alex Osterwaller, founder and CEO of Strategizer. And we've also got Paris Thomas with us today, who is an entrepreneur, a speaker and an advisor. He's co founded a number of startups, including Talent Hacks, where he's reinventing corporate training. And he also works as an executive advisor for Strategizer as well. And I'm Carol Hill, I'm a program director at Strategizer and I'm going to host today's webinar. So welcome. Today, what we're going to be talking about, it's all about AI. We're going to start with the innovation process and how AI can impact that. Then we're going to move to innovation and governance and then we're going to talk a little bit about considerations and a few call to actions. And then at the very end we will have some time for a Q and A.

Thank you very much for the invitation to participate with you and talk about something that's very interesting and cool to me because I'm a tech geek completely and I love talking about technology. But today we'll talk about innovation and AI. What can we do to expedite the process of getting out of the building and talking to customers? But before we go into the whole conversation about innovation, let's align on some terms that are quite important to make sure we're all on the same page, because it feels like AI has been new to our life. But really, I don't know if that's completely true and we need to kind of be on the same frequency. So some definitions, if we start with artificial intelligence, right? I mean, it's kind of cool and sexy, a buzzword. Now, AI, but what is artificial intelligence? It's when machines are simulating human behavior. And if you think about the last couple of years you were talking about, AI is in our life. That's not true. It's been in our life for a very long time. If you think about Apple Siri, that is a form of AI. That's the part where your phone is acting as if it's a human and talking back to you. And then beyond that, one of the other terms or buzzwords we hear is machine learning. And you go, okay, well, if I've heard that term before, it sounds like this kind of crazy stuff that's going on with machines and they're learning all about our world. That must be quite innovative and new. And really, again, it's not. Because what is machine learning? It's an if this, then that statement, if the movie you just watched has also been watched by other people, and those other people then ended up watching these other movies, perhaps these other movies will also be interesting to you. So it's just a bunch of data that's coming together to offer to you what you've seen for over a decade, which is, for example, Netflix recommendations of what movie to watch next. So that's a version of machine learning. And then it gets a bit more interesting because we talk about deep learning. And what is deep learning? Well, it's a subset of machine learning. So when you think of the example I have here is Tesla Autopilot, you have a whole bunch of sensors, like you have cameras, ultrasonic sensors, lidar, all sorts of other things around, gadgets around the car. All of that input is coming in in very interesting ways to create a 3D model of the world around the car. And then that helps the car navigate within that world. So that's sort of a version of deep learning when you're using multimodal ways to create something a bit more complex and find ways to simplify it. We've also heard of nlp, Natural language processing. So Google Translate has been around for so long, but now we all know GPT, right? And so GPT translate for breakfast. Like the fidelity is so much better. But at the same time, understanding and producing human language is not new. It's been around for quite some time. So what's, what's happened that's causing everybody to talk about AI these days? Well, let's continue because one of the biggest manifestations of AI and where we started to pay attention is when AI started to generate content, whether that was text or that's images through software like Dall? E, or if it's music. And that's when it got a little bit, let's admit it, scary, right? Because it started to fill in the gap of where humanity has been really good at doing know so many things for so long. And all of a sudden these machines can generate content. And where it all lands eventually is this other buzzword, if you've heard of LLMs. So let me read this particular definition because this is where it all comes together. An AI model designed to understand, generate and predict text by processing large amounts of natural language data using, using deep learning techniques. So it's every other slide I've shown so far all comes together. That's your GPT, Right. So that's your LLM. And the conversations I keep hearing from customers over and over is that everybody's rushing to create their own LLMs, hosted by their own servers, with their own super knowledge in order to help them get stuff done. What are they trying to get done? Well, let's look at the impact on business, especially if you're new to AI. How is this going to manifest the way we do business? And I'm just going to take some examples from common areas like customer service and marketing and so on and so forth. So imagine we're at the cusp of having 24, 7 customer support with AI agents where you can't tell the difference if the other person on the phone is a real human or an AI agent. And so obviously that's going to create an impact for the business. And you'll see a common thread here. It's cost savings, it's efficiency, it's some form of competitive advantage. Imagine the insane insights you can gather from marketing data and your customers. Data that helps you innovate with new value propositions out there in the marketplace, creating again, an amount of responsiveness in the market we've never seen before. Hence, another competitive advantage. If you're familiar with Excel, one of the early, let's say, versions of automation has been a macro. So if you know what a macro is, it's just another way of saying, Alex and I, for example, we were in these studios that have, I don't know, 100 buttons around us. Every single one of them is a macro that gets 20, 30 jobs done with a single press of a button. So operations wise, you have a ton of ways to create cost savings and efficiency through automated routines that do the donkey work for you. Sales wise, the potential prediction of customer preferences, the ability to be incredibly good at your after sales and improve your customer journey. Again, competitive advantage. Hr. We already have an issue with, you know, a thin talent pool. What if you opened it up to the whole world and could automate the screening process? Again, competitive advantage. Finance. What if you could detect fraud and you know, do some trading automation so while you sleep, you're making money? Again, cost savings, inefficiency, and everything comes together to create this kind of feeling of I need to be cutting edge because if I'm not, somebody else is. So at least I don't know about you. Alex, the clients I talk to, everybody's just rushing to say, I need to understand this. I, I need to get better at it. It, it needs to be faster. It can help me get faster. So I need to get there. But you know, why has everything changed right now? I'll pass it over back to you.

We think AI is new and is not. It's a discipline. Somebody educated me that the discipline itself is actually seven decades old. But it feels very new. Why? Because the accessibility has completely changed. When I did my PhD, actually one of my colleagues was doing a PhD on AI. And that's already like 30 years ago, 20 years ago. What has changed? Let me give you some analogies to give you a feel of why this is so important now and why we actually all need to really pay attention. Ask yourself. Maybe type into the chat window. What big thing happened in 1990? Does anybody remember what happened in 1990? Type into the chat window has nothing to do with AI. This is more broad. Okay, so somebody put World Wide Web. Correct. So Tim Berners Lee at the CERN in Geneva, he launched the first graphical web browser. So why am I mentioning this? Because the Internet per se was not new. But the accessibility completely changed with the graphical web browser. Right? So something changed that made it much more accessible to a much larger group of people. Now let's try 2002. Anybody know what happened in 2002? This one's more difficult, obviously. Again, not related to AI. Yet Google could be year 2K. Well, that's two years later. Okay, 3G. Interesting. Well, Amazon Web Services made parts of their infrastructure accessible to any developer. So why am I showing this? It's because all of a sudden with the credit card, you could access the world's best infrastructure, IT web infrastructure, together with Google and, and Microsoft, and you could use it as a developer to launch your company. You did not need to buy servers, you did not need to understand how to manage that backstage infrastructure. You just got access. So what does that mean? A whole new series of startups could launch without actually having to invest, nor understand the servers because they focused on the front end. So last piece here, 2023, the same thing happened with AI. So OpenAI launched their API officially to Jet GPT GPT 3.5 and that made something accessible with incredible power to everybody. So today any kind of startup can use this stuff. So the competitiveness is going through the roof. So it feels like it's new because there are a lot more companies building stuff around some of these things that are accessible either through large companies. And now we're starting to see more open source stuff and independent models happening as well. So I wanted to show this because it's a game changer. What's possible. Today, I can tell you at Strategizer, we have to talk about AI as well. And how does that change our business? Because they're already the first startups that jump up and, you know, talk about these things and. And what they can do. And that's why it's so exciting to have Paris today on the show, to show what he's been doing in the real world in terms of innovation process, but before.

All right, okay, so let's talk about innovation and the innovation process, and how can that change using AI? Now, before we talk about AI, let's talk about the innovation process itself. Well, the first thing is that you may have seen this before if you're into innovation in very practical ways. But the process of going from an idea to a business we now know can be split up into two different, very distinct mindsets. The part where you're exploring, you're searching, you're testing to ensure that idea has legs before you start building and executing and doing all that costly work. But we also know that there's four distinct phases, phases within that. It's starting with discovery, where you're trying to understand if there's even demand for that particular idea, what we call desirability. And as you try to create some of that evidence, you transition into more sophisticated experimentation that falls in line with validation type of experiments. And as you give yourself that evidence and the confidence, then you make the decision, do I now go into a pilot? And that's what I consider acceleration. Not day one pilot, but much later, when you have evidence, do I go into acceleration and pilot this and start to integrate it and go into a full launch to integrate it as part of my business? But we also know that this whole process isn't really linear, like especially the search and test. It's a bit of a bumpy road, especially if any of you on the call today are entrepreneurs. You know, it's a bit of a bumpy road. And you're probably already thinking that, hey, Paris, it's a lot bumpier than what you're showing us. And you're absolutely right, because this is exactly what the process looks like. I've been through that process multiple times. And let me just tell you that it's not just the process from an operational perspective, it's the emotional perspective. Some of those squiggles go backwards. And so to grapple with that takes a lot of grit and perseverance, but you're doing that fully knowing that you're trying to take the right smart decisions. So you don't break the bank so you don't waste all your money and that you do what we call metered funding. And you're doing that because you're spending more money as you're improving your level of confidence and, and reducing the risk or the idea that is the innovation process. But today I want to focus specifically on this first part, which is where it all begins. Discovery. Lots of ideas coming in. What do we do to discover how do we get out of the building quickly so we can start talking to people. And yes, I do mean really talking to humans, human to human, no AI involvement. I think we're not there yet. We're still in the process of trying to get out of the building to talk to humans. So here's what I kind of roughly sketched out what the process looks like just with some boxes. And what you can see is that sometimes teams come in with ideas, whether it's a corporate team or a startup that has ideas or sometimes they don't have ideas and they need help getting ideas out there or entertained. So we do ideation. We then take a volume of ideas and we walk them through a process we call affinity mapping where we're trying to synthesize these ideas into something more concrete. And eventually we start to create the first prototypes of these ideas, which we call napkin sketches, which are a visual representation of the early ideas that help us align in that process typically takes a day and a half to two days, workshop wise. Right now let's say we have, you know, our top selected ideas. We have a visual representation of what those ideas look like. The next level is what could those ideas look like if they were to become a business. So that's when business modeling comes in. We start to explain how this idea is going to create a business that, that creates, delivers and captures value from the market. Speaking of the market, we'll start to look at the value proposition design, how does it achieve product market fit and do we even have any? And then eventually to be able to deep dive to find answers to those questions. We're going to create customer profiles and customer profiles are going to be very important for us to be able to go out there and co design with our real customers or potential customers what their needs might be. We talk about needs by dividing them into what we call jobs to be done, pains and gains. That whole process takes about two weeks. And then once we're, you know, to be able to actually get out there, we're not out of the building yet. We need to be able to explain very clearly what is the value we're creating for these particular customers explicitly? And the way to do that is using certain assets that we call value prototypes. And I'll show some examples of that in a moment. We then have to create our system of how we're going to conduct qualitative interviews at scale so we can get some evidence. And, and then when we're ready to do that now we can get out of the building. So that whole process of getting all the prep work and the design work, that takes two weeks. So it's not a long time. It's five weeks. Right? Five weeks. You're able to get out of the building and test an idea. Wow, what an incredible fast journey. But it can get faster. So here's where AI comes in. So now this new version, I'm going to take an approach here where we have, we still have the ability to be human and we still have value. So I'm going to illustrate that with the brain. Then we have the AI copilots that are specially designed tools that some of them we've designed some of them, they're out there for you to use, that you can use to copilot and expedite and get there faster. And then when I show the OpenAI logo, I'm speaking specifically about custom GPTs that you can design fit for purpose. Okay. Ideation wise. AI can help with ideation. Like you have no idea. And I'll show examples of that in just a moment. Affinity mapping, I love to still use humans for that because I think there's a process of buy in and alignment that happens between people that that's not there just by reading a screen and output from the AI and then napkin sketching. Easily use AI. The whole process has gone from 2 days to an hour and a half. If we're looking at the business modeling value proposition, customer profiling through custom GPTs and specialized prompts. And I'll show some examples of that. We can condense that down to two hours to be able to create the assets we need. And, and if we're talking about value prototypes, which are again designs and the process of getting out of the building, that whole thing can be condensed to an hour and a half. So we've just gone from five weeks to five hours to have enough assets in our disposal, at our disposal to get out of the building and start testing our ideas. Let's go through specifically examples of how we can do that. So how do we compress this innovation cycle? The first part, as I mentioned before, is ideation. Let Me get here to the left so I don't. Not in the middle of the prompts. So ideation, we can do wider triggering and trigger is when I, when I, when I talk about a trigger, I mean, I mean about what are the things that you ask, you know, a corporate innovation team to help them come up with more ideas. Here is a prompt in the context of my innovation strategy. If you have one, you can upload the PDF. Provide me with seven trigger questions for each where to play area and a where to play area is your specific innovation focus. So you may have multiple where to play areas or you may have one where to play area. You have to define that context and then you say, give me trigger questions so that my team and I can ideate further in this space. So it gives you a whole bunch of very interesting questions that get you to think and triggers. But then you follow up with that prompt and you say, for each of the above trigger questions, come up with an example idea and a visual that represents that idea. Here's an example from a client of mine recently where they were looking specifically at the area of healthy aging. Here's a trigger I came up with within about 30 seconds. And it says, how can we develop tools or services to help older adults manage stress and relax? There's a visual there and that visual is explained by the example, a guided meditation app with exercises specifically designed to seniors. Now, what I found, Alex, personally, is that without the ability to create the visual and the example, and what I was doing before is just the text that I would come up from my head, right? I've gone from people coming up with, you know, 15 to 20 ideas per person that would then put together as a team to 5 to 8x that amount. People are counting their ideas at the end of this session, same amount of time during the session, and they're coming up with 80, 90 ideas each. So maybe it's the visual, maybe it's the, you know, all of it put together. It just gets us triggered better and wider when we start to use some of these tools in that particular way. But let's go beyond that. What happens afterwards? Well, we want to prototype our idea and we want to create some visuals rapidly in order to be able to take our ideas out of the building and show them to customers. So the way we do this is we use a couple of templates. One of them, which is my all time favorite, is called a customer value scene. And in simple terms, what this does, it's basically a cartoon storyboard that allows you to Explain to a customer three things. The moment of need, which is the pain you're trying to solve, how you think the customer is solving that pain today, and with your idea or feature, how the customer might be able to solve it tomorrow. And obviously, tomorrow looks better than today. And so what you do is you use that asset to take it out there and have a conversation in a meaningful way. And we train people on the qualitative interview process and how to get all that data. The second asset that we create, especially when it comes to ideation, is a napkin sketch. And that's just an early manifestation of the idea. It helps the team align on their ideas, park the ones they want to work on later, so on and so forth. So the way we do it, what we've done is we've built our own tool called Rapid Visual AI and we're experimenting with it as we consider integrating it within the Strategizer platform. But for now, I'm going to give you a quick demo so you can see how this can all work together. And you know, Alex, they always say never do a live demo, but here goes nothing. All right, here's Rapid Visual AI. And by the way, as I go through this, just, you know, for you being here, we'll give you a link as a follow up to this webinar that gives you one month free access so you can play with the tool and have some fun. But once you're logged in now in a, in a particular folder, that's the webinar demo Live, I can create customer value scenes, I can create napkin sketches, or I can create logos. And we're coming up with more visual assets that can help discovery teams get there faster. We want to create a customer value scene now. So once you go in there, you type the title of it and we have examples and a sort of a guide. So for convenience, so you're not just watching me type, I'm going to copy paste the stuff that's on the example. So it's a smart home security system. Okay, awesome. What's the next level? Well, describe the visual for the moment of need. So what do I want? Well, I want, I want the image to display somebody who is ready to enter their house, but they're kind of scared to do so because they don't really feel safe. So I explained that in the interest of time, I'm not going to read the whole example, but there we have the example now for today. So how do they deal with that today? Well, today they actually need to install cameras and lights Everywhere and deadbolts. And it's a process that they can't do themselves. But it's just. It's. It's just very cumbersome. So let me just see. Copy too much. There we go. So basically what I'm doing here, I'm describing what do I want the visual to look like so that I can, you know, the same thing I would do when I'm talking to a designer. I want you to design for me something that looks like this. That's exactly what I'm prompting the AI to do here. So I click next and I'm on the final one, which is what does tomorrow look like? Well, tomorrow, obviously, I'm going to have some home automation system turning the lights on as I get close to home, making me just feel safer. So then I click next and then I get a choice of what kind of graphic do I want. Right. So I do a pencil sketchy illustration. Let's go with pencil. Usually they're my favorite. They're very simple, even cartoon. It's awesome. But it does take longer to generate. I get to, you know, check all the different prompts, save and generate. So what's happening now is that you're probably getting very impatient. Why we're generating 15 images as if. Oh, it's done already. It's okay, I'll stop talking. So there we go. You have not only the images already created, but you know what? This is nice. Oh, this is even better. Just go with this one. This one. Now that's kind of creepy. That's not really what I want to say. This one is can do the job. Let me just see. Oh, yeah, yeah, yeah. That's pretty good. Now this one is a bit wonky. Okay, let's go with this one and this one. Okay. They got the phone. Oh, color. Even better. Let's see, let's see. Perfect. This is. This does the job. So I have just created in what minutes, a visual that would have taken me weeks. So I was just working with a client. We did this whole process before. Actually, I wasn't just. I was working with them last year and we didn't have this tool. So they spent three months with multiple calls with a frustrating designer, and they had to create 20 customer value scenes so they can get out of the building and start testing their. Their values explicitly. It cost them a lot of money to do that. And then later, I was having a coffee with the same client. I said, let me show you a tool, and you just wanted to slap me in the face. Why did you bring this to us before. So what happens is I already created a rapid visual before, so you can export that into PDF and then what you do is you go back to your PowerPoint and you add the quote bubbles and here's how you will present it to a customer. I don't feel it's safe to go in with so many break ins in my neighborhood lately. Today I need to put cameras and deadbolts everywhere, but tomorrow, this is amazing. My phone just lets me know everything is safe. I love the feeling of reassurance that I get from my new smartphone installation. So what have I done? I've made the value explicit and obviously this is a starting point for a very rich conversation with your clients or potential clients where you're trying to understand their reaction to this. We haven't built anything, we haven't written a single line of code. We're just testing. And that's what discovery is all about. Obviously you can choose different styles, right? This is another. The same scene, different style, and you can have a lot of fun with it. Within literally half an hour, you've just created something that's absolutely incredible. So then what do we do with the rest of it? So now we have our idea prototypes.

Just one thing, because you know what's interesting, and I want to point this out, Paris is really good at testing. So he was doing this with that exact client, right? And then people started to hand over money because they wanted, they wanted this right away. So you, you can sometimes it's actually very rare that you get that moment of oh my God. And with AI, we will see more value propositions like that. So it's a game changer for the innovation process, but it will also be a game changer for some products where we'll just say, oh my goodness, like that's possible. So, you know, I think it was really interesting how you even tested that and what you got to, which is pretty amazing and obviously that shows the kind of stuff that we need to integrate in the Strategizer platform is if we don't do this, you'll have this connection of disconnected tools, right? So it does change the innovation process substantially.

It does. And you know what I mean? For those of you who are, let's say, more advanced, you're thinking, oh yeah, that's Dall E Mid Journey Stable diffusion. There's all sorts of tools that can get you to visuals. But at the same time, one of the things we did, because we eat our own dog food, we were testing this, right, for enterprise clients. What are the things that matter to them? And one of the things they said, well, I don't want to share my idea with these open source AI tools. I want to be able to share it without feeling that somebody's going to take the idea or the prompt. So obviously we designed something that would delete the prompt as soon as it was put on there on bespoke servers that the client can also host. So it was just understanding what are the things that get a team accelerated, not just here's another tool that you can use. So, okay, I mean there's a couple of other things that we can say here because to be able to get out of the building, not only do you need your idea visuals like we just explained, but you also need to have a really good customer profile which you then co create with the customer. So you have your base, which is all your assumptions built into the customer profile, your jobs, pains and gains. And then you go out there and you have this conversation with the customer and together you validate that customer profile or invalidated with them. But here's a prompt that gets you to create the customer profile. Oh, just a lot faster and better for customer segment. Fill in the gap. Please provide a list of 30 jobs to be done with a good mix of functional social, emotional jobs and make them as specific as you can using metrics. Now this is like part of what we're coaching people to do, but also this prompt allows you to go with this volume of jobs to be done that some of them you might have never considered considered because you're not the customer segment. But obviously the AI would know a little, a few more wider scope things than you would. You do the same for pains and gains. You get there faster and it's okay if it's wrong because the point of this is to get out and talk to humans and validate it. So we found that the process of creating those customer profiles at this stage of innovation just gets a lot faster. The final two, now you're, let's say you're out of the building, you're having conversations, you're documenting all this stuff on the Strategizer platform. And, and what we've built is for every single interview you do, as you're documenting stuff, automatically things are getting aggregated into a designated workspace that allows you to kind of take a step back and look at all the different patterns and insights that are evolving as you try to synthesize all that information and present it for your decisions moving forward. Obviously now there's already An AI component to the Strategizer app that does a lot of interesting things and actually represents a lot of the content of Strategizer. So you can interact with it and in very interesting ways. And finally you get to. A discovery team would end up with the, you know, the most important part of discovery or anything in innovation, which is to share their story. And so what I found very interesting is, and I play with all sorts of different tools day in, day out. There's a particular tool that I use called Beautiful AI and I did a quick little demo here. So I went into this and I said, I want you to design an eight slide pitch deck, which is our format for an innovation Sprint with a title slide, a slide with a napkin sketch, a business model canvas, a customer profile, two slides with idea prototype placeholders, a slide with customer interview feedback, and one with next steps in the strategic why and maybe a conclusion slide. By the way, this is real time. The video is not expedited. There's my slide deck. Now, obviously I need to now go in and put my own content, but I mean, it's a lot better and faster than starting with a blank PowerPoint deck. So ultimately, AI is not a magic bullet, but it can certainly help you get there faster. In some areas, I've experienced a 40x, sometimes 50x multiple of how much faster we can get there. And in some areas, at this point in time, we still need to take the human brain and say, what's your role in this? We need to get out and talk to real customers and learn a bunch of things. But with AI, we can get there faster.

I want to go to just quickly. One aspect that we sometimes kind of forget, it's the organizational side, right? We fall so in love with technology that we forget in order to create real value at scale, we need to change our organizations. I'm just going to give you an example of a company that was pretty good at putting in place governance to embrace value creation from AI. So now we're a little bit away from using AI in the process, but more governance so that we can embrace AI value creation. So Pingan, you know, used to be a traditional banking and insurance conglomerate. These were the areas they were in. Now they said, we're going to invest into a series of technologies. And this was way before Chad GPT became a thing. And AI and brain is. That's the way they formulated it back then with big data, but also cloud. They said, we're going to use this on the one hand to improve existing business models, to Boost efficiency, cut costs, etc. They would hire a ton of PhDs to be able to do this stuff. But what they also did, and this is governance with a strategic vision where you're going to go and how you, how they invest in that. They said we will also invest in completely new arenas that are not directly related to banking or insurance. So they said, or the founder Peter Ma said, we need to become a technology company and not just focus on banking and insurance. So they reallocated 10% of their profits every year into exploration of completely new domains. So what happened? And again this was more much before this became a popular thing across the globe. They came up with a ton of different things and some we might say, well that already shows the ethical aspect that we need to discuss. So to improve the existing, they had some value propositions like an AI based loan application process where you'd apply for a loan through video. Video would analyze, you know, how you're behaving and based on your behavior during that loan application process, you'd get a loan or not. Okay, now clearly, you know, we can have an ethical discussion. Is that even, you know, can we allow for things like that? Business wise, it went through the roof. So their defaults went down and they could improve the loan application process, make everything a lot faster. So pretty crazy. And again, you know, here we can debate, you know, is this something you'd want to do or not? But again, governance of what do we invest in and what's in and what's out. They went beyond that. New value propositions, AI driven products and services. So they built a incredible chatbot again before Chad GPT became a thing they had Ask Bob that today advises 1.2 million doctors. That's a third of the population of doctors in China on a platform with 400 million users called Pingan Good Health. So think of it. A banking and insurance conglomerate said we're going to, that was their governance, their policy. We're going to invest into building Internet companies and only as a secondary kind of thing are we going to be able to sell, in this case insurance services. Wow, that's a pretty strong kind of transformation because they had the governance to invest into completely new ideas. And here's the most extreme. They said we also want to self disrupt in some cases. So instead of just creating AI based banking, they said, well we only have 3% of the banking market, so if we just improve our own bank, just take this thing away here. If we just improve our own bank, not good enough. What if we used all of these services that we use in our own banking and insurance. And we offered this to all Asian clients first and then globally. So they built a banking platform for other banks and attack their own business. Think of it now, which other company do you know that this did this in a different domain? Amazon, with Amazon Marketplace. So today anybody can sell stuff that's cheaper than the stuff that you buy from Amazon because they built the infrastructure and they earn from the infrastructure. So I just wanted to show this because the governance aspect is super important so that we can benefit not just from improving. So Paris already mentioned, most of our energy goes into improving the existing. But when the existing is dying, you want to invent the future. So Pingan with one connect their banking with a good doctor. They really kind of broke through. But here's the thing in governance also they said to invent new things, we need to be willing to fail. So they had pretty big failures in the secondhand car platform, real estate platform. They died. But they knew that without taking some risk they cannot innovate at scale. And then they actually made better acquisitions based on what they learned. So Ping an grew from the top 500 companies in the world to top 30, top 20 depending on the year you look at because they had a governance that embraced this kind of value creation. So Jessica, who was CEO, co CEO for a long time, built this entire value creation system with the right governance.

So let's go on to considerations and call to actions and then we'll move into our Q and A. So a little bit of doom and gloom for the very end and that is thinking about regulations and other things that we should be mindful of when using AI. So when we're thinking of that, you're probably thinking about ethics. One key thing we know is that humans are biased and humans are the ones that are making a lot of the decisions around what data models and data sets we should be using and how we should be putting our algorithms together and all of that. So what do we know about AI and it's. And the bias that we can see in it? A few of you have commented on it. We have an example from Mid journey. I'm sure you've seen all sorts of pictures with people with six fingers or legs that are winding about the chairs and, and all sorts of things. But there's a, a study that was done not too long ago where they were asking midjourney to create some images of people in specialized professions. And really brilliantly they included younger people and older people, which is great for those of us who are getting on. But what did they do? Of course they. All of the older people were always men, so all the females kind of disappeared from that, reinforcing that gender bias of women in the workplace. So that's one thing to consider. Another is risk to ip. And there's two sides of that. It's ensuring that we know how to keep our own IP safe, but also thinking about what kind of IP and copyright is out there with the data sets that we are using. It's not always clear. And there was an example not too long ago, just last December, with the New York Times and some other newspapers suing OpenAI and Microsoft, saying that they used their material without approval to train their chatbots and which is now being used as a news outlet and a source of reliable information that is competing with those news agencies. So I haven't heard about the results of that, but thinking about regulation, I'm sure there will be some more things there coming out of that and then a third one is all around quality and, you know, ensuring that we are happy with the quality that the AI is producing. So there's a simple example of a newspaper Gannett in the US who was using an AI tool called Lead AI to create some newspaper articles automatically. And they were really widely mocked on social media because they were poorly written, they lacked details that made the newspaper articles really nonsensical, they're poorly written, repetitive, etc. Etc. So they've had to pause the use of that. And then the fourth consideration is really the impact that decisions that are made based on AI have on, on humans. And so there's loads of examples. One of them is unfortunate. Around 2019, 20, 2020, during COVID time there were some algorithms that were being used in healthcare for insurance, trying to figure out who needs to, who should be prioritized for care, who has the highest need. And it was actually flagging one race over the other, which is really unfortunate. So those are all things that need to be considered. So just a question for you in chat real quickly. Do the risks associated with AI outweigh its benefits? And you can just say yes or no, what do you think? But we have a quote here from Emma R. Camp Bloom, who is AI ethics researcher, chair of UNESCO's World Commission on Ethics and Scientific Knowledge, member of UN AI Advisory Board, etc. Only sustainable AI is lawful AI, trustworthy AI and responsible AI. But she also says it has nothing to do with stopping innovation. So we need to continue to, to use it as a, as a tool. So some Call to actions for you. Creating an AI task force. All good task force and project teams have a purpose and a goal. Also a diverse and skilled team group of team members, always diversity is best for teams. We know that with innovation as well, but especially when you are making decisions around AI, you don't want to bring one subset of your bias with you. You want a diverse team as much as possible. Setting those expectations and accountability and transparent communication so your teams know what's happening and also your customers know how their data is being used is always brilliant. And then also ensuring that you know the efficacy and the impact of your use of AI. And then the second one is really thinking about how can you use AI responsibly. And there's loads of frameworks out there and I'm listing a few. We're not endorsing any one of these, but we're just listing a few to get an idea of what kinds of frameworks are out there for responsibility, responsible use a lot of key threads here with transparency, accountability, responsibility, security. So make sure you're aware of what those frameworks are and keeping up with the regulations as well. And then finally, be like a cat, Stay curious, know where your data is coming from, sense, check it with humans, and just keep up to date with regulations. So finally, artificial intelligence is not a substitute for human intelligence. It's a tool to amplify human creativity and ingenuity. And I think from what we've seen with Paris and Alex as well, it can increase the speed that you go at with innovation.